$250K in 12 Days: Full Autopsy of Polymarket's #3 Bot — Every Trade, Every Signal

Final Numbers (After Full Resolution)

Where the Money Actually Came From

After fully resolving all trades via the CLOB API, the picture is decisive. The profit is not spread evenly — it is brutally concentrated:

The 12 Jackpot Bets — Full List

There were exactly 12 bets at prices below 2¢ (50×+ odds) across the entire month. No small ones — every jackpot bet was at least $148. Here they all are:

Notice something critical: the bot never placed a small jackpot bet. There are no $10 or $50 bets at 1¢. When the jackpot signal fires, the bot goes in with $148–$936. When the signal doesn't fire, it doesn't bet at 1¢ at all. This isn't guessing — it's precise sizing.

Kelly Criterion: Why $311 and Not $50

This is the question everyone asks. The answer is a concept called Kelly Criterion — the mathematical formula for how much to bet when you have an edge.

Imagine this: you find a coin that lands heads 60% of the time, but everyone else thinks it's 50/50. For every dollar you bet, you win $1 if heads, lose $1 if tails. How much of your money should you bet each flip?

• Bet everything? → One bad run wipes you out.
• Bet $1 always? → Safe, but grows too slowly.
• Kelly says: bet 20% of your bankroll (60%−40% = 20% edge).

The match between theoretical Kelly sizing ($300) and actual average bet ($311) is not a coincidence. This bot was built by someone who understands bet sizing mathematics at a professional level.

By Asset: ETH Dominated

ETH generates the most profit despite having fewer trades than BTC. That April 10 XRP jackpot (162×) and the April 20 ETH cluster drove this. The bot is opportunistic — it follows wherever the mispricings appear.

April 10: One Full Day, Every Trade

To understand exactly how this bot operates, we reconstructed every one of its 334 trades on April 10 — its most profitable single day before April 20. Total PnL: +$43,317.

Summary by strategy type

The two moments that made the day

Between these three moments, the bot was running its background strategy: placing LOW and HIGH hedge bets every 5–15 minutes across BTC, ETH, and XRP markets. All of those together generated +$67 net — less than 0.2% of the day's profit.

Key April 10 observations

• The bot entered every single 5-minute market within 18–30 seconds of opening. It never waited to see which direction crypto moved first.
• It placed two bets per market simultaneously — one LOW (potential jackpot), one HIGH (hedge). The HIGH bet slightly reduces the net cost of the lottery ticket.
• The jackpot bets appeared at 09:35 and 10:00 UTC — during what appears to have been a period of unusual volatility in ETH and XRP. The bot's model detected that the market was underpricing extreme moves.
• Before and after the jackpots, the bot continued its normal rhythm — hundreds of small bets, barely profitable in aggregate. This is not a coincidence: the normal bets generate data and minor income while waiting for the model to fire.

The Core Strategy, Explained Simply

Think of the bot as someone who sells insurance while also buying lottery tickets:

1. The hedge bet (HIGH side, >50¢): "I think BTC/ETH will keep going in the current direction." You collect a small premium — $5–15 — if you're right. This is the insurance-selling part.
2. The lottery bet (LOW side, <10¢): "I think there's a chance of an extreme reversal that the market is ignoring." You pay a small amount to win big if the extreme move happens. This is the lottery-ticket part.
3. When the model detects a major mispricing (market says 1% probability, model says 3–5%), it sizes up the lottery bet to Kelly-optimal size: roughly 2% of its estimated bankroll.

Why Did the Bot Stop on April 21?

We don't know. The last recorded trade was April 21 at 14:56 UTC. Possible explanations:

• Profit-taking: After generating $250K in 12 days, the operator may have paused to withdraw and reassess.
• Model recalibration: After the April 20 cluster (multiple wins and losses on the same day), the model may have entered a review/retrain period.
• Capital constraints: Some jackpot bets on April 20 lost ($482 BTC loss, $207 ETH loss). The operator may have needed to replenish capital.
• Market structure change: Polymarket may have adjusted liquidity or pricing on extreme-odds markets, eliminating the edge temporarily.

We have a Telegram monitor running that will alert the moment this wallet places a new BTC trade. @Polylenspro_bot subscribers will be notified in real time.

Can This Be Replicated?

The bot is inactive, so direct copy-trading isn't currently possible. But the principle can be studied and potentially rebuilt:

What you'd need to replicate the strategy

1. A volatility model: Something that calculates whether the market is underpricing extreme crypto moves. The bot uses proprietary logic — but the underlying math (normal distribution, historical σ) is publicly available.
2. Real-time price data: BTC/ETH/XRP price feeds to detect when a 15-minute window is developing unusual momentum or volatility.
3. Kelly sizing: The formula is simple once you have a reliable edge estimate. The difficult part is knowing your true edge — getting that wrong destroys the strategy.
4. Speed: The bot enters within 18–30 seconds of market open. Manual execution is probably too slow for the jackpot bets — though the hedge bets can be placed manually.

Track This Wallet Live

Live wallet analysis, trade history, and win rate breakdown:

PolyLens Wallet Analysis: 0x7347d329…d35 →

For alerts when this wallet activates again: @Polylenspro_bot

See also: #1 Wallet ($668K BTC bot) · #2 Wallet ($286K ETH/SOL)